Peng-shuai Wang

Assistant Professor at Peking University

Haidian District, Beijing, China
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts

Summary

👤
Senior
🎓
Top School
Peng-shuai Wang is an assistant professor at Peking University with 12 years of experience bridging deep learning, geometry processing, and computer graphics. After earning a Ph.D. in Computer Graphics from Tsinghua University, he spent four years as a researcher at Microsoft Research Asia, contributing to production-quality 3D learning tools. His open-source work includes core contributions to the widely cited microsoft/O-CNN project, where he enhanced octree convolution layers and tooling for 3D shape analysis and retrieval experiments. Combining rigorous academic research with hands-on ML engineering, he focuses on scalable representations and algorithms for 3D data. Based in Beijing’s Haidian District, he maintains an active research homepage that highlights ongoing projects and collaborations.
code12 years of coding experience
job4 years of employment as a software developer
bookPh.D, Computer Graphics, Ph.D, Computer Graphics at Tsinghua University
github-logo-circle

Github Skills (6)

mask-rcnn10
faster-rcnn10
convolutional-neural-networks10
caffe10
octree10
python5

Programming languages (6)

TypeScriptC++CTeXPythonCuda

Github contributions (5)

github-logo-circle
microsoft/O-CNN

Jun 2017 - Apr 2022

O-CNN: Octree-based Convolutional Neural Networks for 3D Shape Analysis
Role in this project:
userBack-end Developer & ML Engineer
Contributions:2 releases, 145 commits, 12 PRs in 4 years 11 months
Contributions summary:Peng-shuai contributed to the core structure of the project by updating the caffe proto files, which are fundamental for neural network architecture definition. They also made significant changes to the octree layers, including modifications to the octree base convolution, unpooling, pooling, and full voxel layers, indicating a focus on the 3D shape analysis pipeline. Further, they added details about the retrieval experiment and improved the tools feature_pooling.cpp & caffe.cpp, suggesting active involvement in model development, experimentation, and tooling.
pytorcho-cnndeep-learningcomputer-visionconvolutional
wang-ps/octree-ext

Oct 2019 - Nov 2022

Contributions:8 commits, 13 pushes in 3 years 1 month
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.
Request Free Trial